Michał Horzela

𝗥𝗼𝗹𝗲 𝗕𝗮𝘀𝗲𝗱 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 𝘃𝘀. 𝗔𝘁𝘁𝗿𝗶𝗯𝘂𝘁𝗲 𝗕𝗮𝘀𝗲𝗱 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 Authentication = Who are you? Authorization = What are you allowed to do? There are two authorization patterns you must know: (NOTE: We recognize that scope-based authorization is also a pattern. However, it's specific to the context of APIs and delegated access, and nowhere as common as the two below) Role-Based Access Control (RBAC) ‣ Users are assigned roles (such as admin, editor, or viewer), and each role has specific permissions. ‣ Simple to implement and works well for most business applications. ‣ In a Google Doc, users can be assigned roles like "Viewer" (can only view files), "Commenter" (can view and comment), or "Editor" (can view, comment, and edit files). The API checks the user's role before allowing actions like editing or deleting a file. If three new people need access to the doc, instead of giving them access each, they can just be assigned roles, through which they automatically get assigned the access level related to that role. If there is ever a new feature that needs to be added to an application, and we want specific roles to be able to use it - RBAC makes it easier. Attribute-Based Access Control (ABAC) ‣ Grants access based on attributes. There are four main categories of attributes: ‣ User (or subject) attributes "Who are you?" ‣ Resource (or object) attributes "What do you want to access?" ‣ Action attributes "What do you want to do with it?" ‣ Environment attributes "When, where, and how is the request being made? For example, we may have a policy that says, "employees can only share documents with external business during business hours". The API checks user and resource attributes before granting access. If the employee tries to share documents outside of business hours, when the IT admin staff may not be actively monitoring, the system will deny permission. Which one is better? It all depends on the business needs. But at the end of day, it doesn't need to be either or. A lot of organizations use a hybrid approach. For example, you may use the roles to assign the attributes. Though, you can make both as simple, or as complex as possible. You should pick whichever one meets the organization's needs, not because you deem that one is "better than the other". --- Join System Design School -> https://lnkd.in/eNmimeaQ

71

⚡ New GitLab Duo Agent Platform Flow: Issue ➡️ MR in minutes No more hunting through codebases for simple UI tweaks. 1️⃣ Create issue with clear requirements 2️⃣ Click "Generate MR with Duo" 3️⃣ Review proposed changes 4️⃣ Merge A way for you to spend less time on setup and more time building! 🙌

131

7 Kommentare

🚧 From monolithic pain to microfrontend gain. At Druva, we faced the classic scale problem: → Slow builds → Complex test cycles → Coupled components that broke each other So, we built a multi-level microfrontend architecture — with shared utilities, reusable core components, and workload-specific UI layers. The result? ⚡️ 5x faster builds ✅ 50% boost in dev velocity 📦 Scalable code reuse across teams Curious how it all works? https://ow.ly/5iN050WAcMN

43

2 Kommentare

𝐀𝐳𝐮𝐫𝐞 𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧 𝐃𝐞𝐩𝐥𝐨𝐲𝐦𝐞𝐧𝐭 𝐀𝐫𝐜𝐡𝐢𝐭𝐞𝐜𝐭𝐮𝐫𝐞❗️ Ship features to production AKS while keeping every commit traceable, every secret safe, and every cluster consistent. This blueprint unifies Dev ➜ QA ➜ Prod with GitHub‑first CI/CD and Azure‑native guardrails 👇 ✅ 𝗕𝗿𝗶𝗱𝗴𝗲 𝘁𝗼 𝗞𝘂𝗯𝗲𝗿𝗻𝗲𝘁𝗲𝘀 – hot‑reload microservices running locally against a live AKS cluster.  ✅ 𝗩𝗶𝘀𝘂𝗮𝗹 𝗦𝘁𝘂𝗱𝗶𝗼 𝗖𝗼𝗱𝗲 + 𝗚𝗶𝘁𝗛𝘂𝗯 𝗥𝗲𝗽𝗼𝘀 – inner‑loop coding with integrated issues, PRs, and tests.  ✅ 𝗚𝗶𝘁𝗛𝘂𝗯 𝗔𝗰𝘁𝗶𝗼𝗻𝘀 – container‑build & unit‑test pipeline feeding the DevTest Container Registry.  ✅ 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁 𝗔𝗞𝗦 – v2‑committed services iterate fast against isolated Cosmos DB.  ✅ 𝗤𝗔 𝗔𝗞𝗦 – promotion gate for release‑candidate images, smoke & load tests.  ✅ 𝗔𝘇𝘂𝗿𝗲 𝗕𝗼𝗮𝗿𝗱𝘀 – work‑item tracking that auto‑links commits, builds, and deployments.  ✅ 𝗔𝘇𝘂𝗿𝗲 𝗣𝗶𝗽𝗲𝗹𝗶𝗻𝗲𝘀 – delivers signed images from DevTest ACR to Production ACR with policy checks.  ✅ 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗔𝗞𝗦 – version‑pinned workloads, horizontal pod autoscaling, and blue‑green swap support.  ✅ 𝗔𝘇𝘂𝗿𝗲 𝗞𝗲𝘆 𝗩𝗮𝘂𝗹𝘁 & 𝗔𝘇𝘂𝗿𝗲 𝗔𝗰𝘁𝗶𝘃𝗲 𝗗𝗶𝗿𝗲𝗰𝘁𝗼𝗿𝘆 – centrally managed secrets and identity across all stages.  ✅ 𝗔𝘇𝘂𝗿𝗲 𝗠𝗼𝗻𝗶𝘁𝗼𝗿 – unified logs, metrics, and alerts wired into every cluster. ️𝐋𝐢𝐟𝐞𝐜𝐲𝐜𝐥𝐞 ♻ 1️⃣ Code: Dev pushes to GitHub → Actions build, test, and push to DevTest ACR.  2️⃣ Validate: Dev & QA AKS pull images; automated tests + manual sign‑off gate promotion.  3️⃣ Release: Azure Pipelines signs artifacts, updates Prod ACR, and triggers rolling upgrade on Prod AKS.  4️⃣ Operate: Key Vault injects secrets; Monitor streams telemetry; Boards closes the loop with traceability. Visualise this end‑to‑end flow in seconds with the Cloudairy Azure Application Deployment template - https://lnkd.in/dmEag_Uh  #Cloudairy #Cloudchart #Azure #AzureDeployment #CloudArchitecture #DevOps

101

12 Kommentare

How to write structured prompts: Use the JSON format to get better results. Unstructured prompts force the model to guess: • How long should this be • What sections to include • How much detail to add • When to stop Instead, here's a template to copy-paste: {  "task": "describe the job here",  "input": "your text, data, or image prompt goes here",  "options": {   "setting_one": "value",   "setting_two": "value"  },  "output_format": {   "type": "how you want the reply",   "extra_details": ["list", "any", "extras"]  } } To go much deeper, go to → how-to-prompt.ai

170

25 Kommentare

Async Rust’s disk I/O is not always ideal, largely because fully asynchronous file I/O APIs are relatively new on most mainstream operating systems. #P99CONF speaker Tzu Gwo explains how they developed Fusio to provide a minimal cost abstraction layer that delivers unified asynchronous data access across various platforms. https://lnkd.in/eDi-gpX7 #ScyllaDB

32

2 Kommentare

📈 We released DuckDB 1.4.0 today. This is our first LTS release that comes with one year of community support. 👉 The new release packs many new features for security, data lake support and productivity: swipe below for a quick summary. ⏱️ For a more complete list of performance optimizations, distribution improvements and new integrations, see the announcement blog post – link in the comments.

742

33 Kommentare

With Great Access Comes Great Responsibility! Access tokens are powerful - too powerful, if you don’t rein them in. Developer Advocate Andrea Chiarelli's deep dive into OAuth 2.0 explains how to keep those tokens on a tight leash using the Principle of Least Privilege. 👀 Learn how to: ✔️ Scope like a pro (no more scope=*) ✔️ Design granular permissions (e.g., orders:read, not just orders) ✔️ Validate tokens the right way, not just checking if they’re expired ‼️ Bonus: It includes a checklist you can actually use before you ship. 🔗 https://lnkd.in/e9Jacend

63

2 Kommentare

⏩ Understanding Application Architectures → Monoliths: The Original Giants • One big, tightly connected application - like a complex machine with all parts linked • Simple to develop and deploy at first • Hard to scale, risky to update, and a single failure can bring everything down → Microservices: The Flexible Builders • Independent service “bricks” working together but standing alone • Easy to scale and update individual components • Needs complex infrastructure and team coordination → Modular Monoliths: The Balanced Middle • A monolith designed with loosely coupled internal modules • Combines easier scaling than traditional monoliths with simpler management than microservices • Demands careful planning for true modularity → Distributed Monoliths: The Hidden Trap • Looks like microservices but acts like a tightly coupled monolith • Faces scaling problems and complexity of distribution • Often results from premature optimization or flawed design → Every architecture has pros and cons. → Choosing the right one depends on your project’s goals and constraints. → More differences are covered in post credit:-Ashish Joshi follow:-Tech In Nutshell

35

8 Kommentare